Association of Life’s Essential 8 and Life’s Simple 7 with metabolic-associated fatty liver disease in the United States

Abstract

BACKGROUND

Cardiovascular health (CVH) plays a crucial role in overall health, but its association with metabolic-associated fatty liver disease (MAFLD) remains unclear.

AIM

To investigate the relationship between CVH, measured using Life’s Essential 8 (LE8) and Life’s Simple 7 (LS7), and the prevalence of MAFLD.

METHODS

This cross-sectional study had a sample of 2234 individuals, representing approximately 120 million individuals in the United States. Baseline parameters were compared between the LE8 and LS7 groups. Logistic regression models were used to evaluate the relationship between LE8, LS7, and MAFLD, while taking into account confounding factors. The investigation employed restricted cubic splines to investigate non-linear associations. Subgroup analyses and sensitivity studies were performed to evaluate the strength and reliability of the results.

RESULTS

Higher LE8 and LS7 scores were significantly associated with a decreased risk of MAFLD, even after controlling for demographic, socioeconomic, and clinical variables. This association demonstrated a non-linear pattern, with the most dramatic risk reduction observed at higher CVH levels. Individual CVH components, notably healthy behaviors and factors, exhibited strong relationships with MAFLD. Subgroup analyses indicated consistent relationships across several demographics. Sensitivity tests utilizing other MAFLD definitions validated the robustness of the findings.

CONCLUSION

Higher adherence to CVH criteria, as indicated by LE8 and LS7 scores, is associated with a significantly lower risk of MAFLD. These results emphasize the need to advance CVH to control and avoid MAFLD.

Key Words: Cardiovascular health; Life’s Essential 8; Life’s Simple 7; Metabolic-associated fatty liver disease; National Health And Nutrition Examination Survey

Core Tip: Higher adherence to cardiovascular health (CVH) metrics, assessed using Life’s Essential 8 and Life’s Simple 7, is strongly associated with a reduced risk of metabolic-associated fatty liver disease (MAFLD). This large cross-sectional study demonstrates a non-linear relationship, with the most significant risk reduction observed at higher CVH levels, emphasizing the importance of promoting CVH for the prevention and management of MAFLD.

INTRODUCTION

Metabolic-associated fatty liver disease (MAFLD), formerly known as non-alcoholic fatty liver disease (NAFLD), is an increasing global health concern[1]. With an estimated 25%–38% of the world’s population affected, the rising incidence of MAFLD is alarming, as it significantly reduces survival rates and poses a considerable threat to public health[2]. The perils of MAFLD are multifaceted, extending beyond liver-related complications. It is a leading cause of advanced liver pathologies, including cirrhosis and hepatocellular carcinoma, which contribute to a substantial number of mortalities annually[3]. The limitations in treatment options, particularly for end-stage liver disease, highlight an urgent need for novel therapeutic approaches. Furthermore, MAFLD is closely associated with type 2 diabetes and obesity, significantly increasing the risk of cardiovascular diseases, metabolic disorders, and various forms of cancer, thereby increasing the burden on healthcare systems and affecting the quality of life for millions of individuals worldwide[4]. Recognizing the complexity of MAFLD, the identification of new indicators that predict the onset and progression of fatty liver disease is of paramount importance. The understanding that not all individuals with obesity develop MAFLD suggests that other factors may play a role in its pathogenesis.

MAFLD has been increasingly recognized for its strong association with cardiovascular diseases[5-7]. Studies have shown that higher Life’s Simple 7 (LS7) scores are associated with a lower prevalence of NAFLD[8,9], suggesting that LS7 not only serves as a cardiovascular health (CVH) metric but may also be instrumental in the early identification and prevention of liver diseases such as MAFLD. In 2022, the American Heart Association (AHA) expanded upon LS7 by introducing Life’s Essential 8 (LE8), which incorporates sleep health as an additional parameter. LE8 has since demonstrated superior predictive capabilities over LS7 for various cardiovascular outcomes[10]. However, despite the evident advantages of LE8, there remains a gap in research concerning its relationship with MAFLD. It is noteworthy that few studies have investigated the applicability of LE8 to the revised MAFLD nomenclature; understanding this relationship might help to clarify the preventative and therapeutic approaches for liver disease and CVH.

In this study, we investigate the relationship between LE8, LS7, and MAFLD, hypothesizing that both scoring systems may have a major impact on the onset and course of MAFLD, especially in the United States population. We aim to determine the strength of these correlations using data from 9254 adults from the National Health and Nutrition Examination Survey (NHANES) 2017–2018, therefore providing a more sophisticated clinical screening and diagnosis for MAFLD and associated metabolic diseases. Furthermore, this study will investigate the relationships of LE8 and LS7 with various definitions of non-alcoholic fatty liver and will include a larger number of significant factors compared to previous research[11,12].

MATERIALS AND METHODS

Study population

The NHANES study collected data on the dietary habits and general health of Americans, including both adults and children. The National Center for Health Statistics Research Ethics Review Board approved the survey, which used a sophisticated sampling technique. One important addition to the 2017–2018 study was the use of ultrasonography and vibration-controlled transient elastography for liver evaluation. A total of 9000 adults participated in this phase of the study. Individuals aged under 20 years, pregnant women, those ineligible for or with incomplete elastography examinations, and those lacking data on the Dietary Approaches to Stop Hypertension (DASH) were excluded (Figure 1).

Figure 1  Flowchart of sample selection.

Definition of MAFLD

This study evaluated liver fibrosis and steatosis using the median controlled attenuation parameter (CAP) and liver stiffness measurement methods. A CAP threshold of 285 dB/m was set along with one of three criteria to detect hepatic steatosis: having type 2 diabetes mellitus (T2DM), being overweight or obese (body mass index [BMI] ≥ 25 kg/m²), or having a BMI < 25 kg/m² but exhibiting at least two metabolic risk abnormalities. These metabolic risk abnormalities included low-density lipoprotein-cholesterol levels below 40 mg/dL for men and 50 mg/dL for women, prediabetes indicators (fasting glucose of 100–125 mg/dL or hemoglobin A1c levels of 5.7%–6.4%), a homeostasis model assessment of insulin resistance score of 2.5 or higher, and plasma high-sensitivity C-reactive protein (CRP) levels exceeding 2 mg/L[13].

Steatotic liver disease (SLD) was detected using CAP scores equal to or exceeding 263 dB/m (≥ S1), ensuring a sensitivity cut-off of 90%[14]. The concept of MAFLD was defined as the presence of SLD with one or more cardiometabolic criteria[15]. Significant fibrosis (≥ F2) was characterized by liver stiffness values equivalent to or above 8 kPa in instances where there were no other underlying liver disorders, substantial alcohol consumption, or the use of steatogenic medications[16].

A CAP score of ≥ 285 dB/m was established as diagnostic of a high risk of NAFLD, whereas advanced fibrosis (F3-4) was characterized by liver stiffness measurements of ≥ 8.6 kPa, coupled with an interquartile range/median (IQR/median) of liver stiffness below 30%[17,18].

Definitions of LE8 and LS7 metrics

LE8 was introduced by the AHA as an expansion of the LS7 paradigm, comprising four health behaviors (diet, physical activity [PA], nicotine exposure, and sleep health) and four health metrics (BMI, blood lipids, blood glucose, and blood pressure [BP])[19,20]. These characteristics were assessed at baseline to determine LE8 scores. Dietary patterns were evaluated using the DASH technique, conducted via face-to-face 24-hour dietary recall interviews employing the USDA Automated Multiple-Pass technique to quantify food and beverage intake. PA included self-reported time spent on activities such as walking, bicycling, domestic duties, and leisure interests during the week[21].

The LE8 score was assessed on a scale of 0–100, with the final score calculated as the unweighted average of all indicators. The entire LE8 score was then categorized into three levels: high (80–100), moderate (50–79), and low (0–49). Detailed directions for applying the LE8 scoring algorithms to NHANES data for adults were included in the online materials and presidential advisory[22].

LS7, the precursor to LE8, excluded sleep health and employed a simpler scoring algorithm. Each LS7 metric was graded into poor, moderate, or ideal levels, corresponding to scores of 0, 1, and 2, respectively. The aggregate LS7 score was categorized as insufficient (0–7), average (8–10), or optimum (11–14), as indicated in earlier studies. Comprehensive instructions for applying LS7 scoring algorithms to NHANES data for adults were provided in the online materials[23].

Covariates

In our study, confounders such as age, sex, race/ethnicity, education, poverty-income ratio, marital status, disease history, alcohol consumption, sleep disorders, and depression were considered based on prior research. The NHANES database provided important covariate data, including BMI, gamma-glutamyl transferase (GGT)(IU/L), and CRP (mg/L).

Age at baseline was obtained using participants’ birthdates and baseline assessment dates and then classified into three groups: < 40 years, 40–64 years, and ≥ 65 years. Self-reported race/ethnicity was categorized as Mexican American, non-Hispanic Black, Non-Hispanic White, Other Hispanic, or Other. Educational level was classified as less than high school, high school or equivalent, and above high school. The poverty-to-income ratio (PIR) was divided into three categories: ≤ 1.30, 1.31–3.50, and > 3.50. Participants scoring ≥ 10 on the Patient Health Questionnaire-9 (PHQ-9) were categorized as suffering from depression. Marital status was categorized as married (or living with a partner) or single (or widowed/divorced/separated), while smoking status was categorized as current smoker, past smoker, and never smoker. Participants were defined as having comorbid conditions if they reported at least one of the following medical conditions: diabetes, kidney failure, kidney stones, heart failure, stroke, hepatopathy, rheumatoid arthritis, or cancer. Information on sleep disturbance was gathered through self-reported responses. Comorbid conditions were analyzed as categorical variables (yes/no).

Statistical analysis

Owing to the sophisticated four-stage sampling strategy adopted by NHANES, the WTDRD1 weights were produced by altering the Mobile Examination Center sample weights (WTMEC2YR) to account for extra non-response and unequal distribution of food intake data collection by day of the week. Weighted univariate logistic regression analysis was employed to filter covariates and identify relevant factors. The strata (SDMVSTRA) and main sampling units (SDMVPSU) were then used to reconstruct the data to reflect the entire non-institutionalized civilian population of the United States[24].

Continuous variables having a normal distribution were provided as mean ± SD, whereas skewed continuous variables were characterized using the median and IQR. Categorical variables were presented as frequencies and percentages. Statistical analysis included the χ² test or Fisher’s exact test for categorical data, one-way analysis of variance for normally distributed continuous variables, and the Kruskal-Wallis H test for skewed distributed variables to analyze differences across groups. To address multiple comparisons, modifications to P-values were performed using the Bonferroni correction, Tukey, or LSD techniques, as applicable. Univariate and multivariate binary logistic regression models were applied to evaluate the associations between LE8, LS7, and MAFLD. In these models, LE8 and LS7 were regarded as categorical variables indicative of clinical diagnoses. Confounders were selected based on clinical significance and existing scientific evidence. Three models were constructed: Model 1 accounted for age and sex; Model 2 further corrected for ethnicity, education, marital status, PIR, and alcohol consumption status; and Model 3 also included comorbidities.

This technique attempted to uncover possible nonlinear interactions. To evaluate possible nonlinear dose-response relationships between LE8, LS7, and MAFLD, we applied a limited cubic spline model to construct smooth curves. In this model, LE8 and LS7 were handled as continuous variables, with four knots positioned at the 5^th, 35^th, 65^th, and 95^th percentiles, following Harrell’s recommendations. Nonlinearity was tested using a likelihood ratio test, comparing a model having just linear terms to one integrating both linear and cubic spline components. After examining the smoothing curves, we proceeded to develop a two-piecewise linear regression model to discover any threshold effects while adjusting for confounding factors. Subgroup analyses were carried out based on subgroup characteristics, with interactions between subgroups investigated by a likelihood ratio test. No covariate exhibited missing values exceeding 5%. Furthermore, the application of K-Nearest Neighbors imputation to address missing data in all covariates did not result in substantial changes in the study outcomes. Sensitivity assessments included complete-case analysis and alternative definitions of outcome variables, with findings consistent across these approaches, as stated in the Supplementary material.

All statistical analyses were conducted using R Statistical Software (Version 4.2.2) and the Free Statistics Analysis Platform (Version 1.9.2, Beijing, China). Statistical significance was determined at a two-sided P value < 0.05.

RESULTS

Baseline characteristics

The fundamental characteristics of the participants are presented in Table 1, categorized based on their adherence to LE8. After undergoing comprehensive screening following pre-established inclusion and exclusion criteria, a total of 2234 patients were included in the study, which is representative of approximately 120 million individuals in the United States. The weighted prevalence of MAFLD in the study population was 40.51%. The distribution of a number of clinical and demographic factors varied significantly across the LE8 groups. Interestingly, gender distribution differed significantly (P = 0.0182) with a larger percentage of men in the low LE8 group. The average age was highest in the moderate LE8 group and lowest in the high LE8 group, indicating a significant difference (P < 0.0001). In addition, socioeconomic variables such as PIR (P = 0.0005) and educational level (P = 0.0010) showed significant differences across LE8 groups.

Table 1 Weighted characteristics of the United States adult population based on NHANES 2017-2018, percentage (%)/mean ± SD/median (25^th-75^th percentiles).

Variables	LE8
	Total	Low	Moderate	High	P value
N	2234	365	1411	458
Weighted N	126883299.26	17279315.83	77128139.71	32475843.72
Gender
Male	47.13	52.50	49.46	38.71	0.0182
Female	52.87	47.50	50.54	61.29
Age	46.532 ± 16.63	53.419 ± 15.20	48.173 ± 16.49	38.972 ± 14.93	< 0.0001
Age group (yr)
< 40	32.00	27.78	31.99	34.28	0.6517
40-64	43.96	46.96	44.37	41.37
≥ 65	24.05	25.27	23.64	24.35
Poverty-to-income ratio	3.22 ± 1.62	2.71 ± 1.63	3.20 ± 1.60	3.55 ± 1.58	0.0005
Education
Less than high school	1.83	1.64	2.14	1.22	0.0010
Completed high school	5.58	11.11	5.32	3.24
Beyond high school	92.59	87.25	92.54	95.54
Marital
Single	62.86	63.24	64.58	58.55	0.0844
Married	37.14	36.76	35.42	41.45
Race
Mexican Americans	8.31	9.84	8.87	6.16	0.1569
Non-Hispanic Blacks	6.36	4.97	6.22	7.42
Non-Hispanic Whites	64.80	63.38	64.62	66.00
Other Hispanics	10.12	12.74	10.60	7.59
Other races	10.41	9.07	9.69	12.83
Smoke
Never	61.58	28.14	58.75	86.07	< 0.0001
Former	23.39	35.72	25.27	12.38
Current	15.03	36.14	15.98	1.55
Alcohol
No	79.03	77.50	80.09	77.33	0.6295
Yes	20.97	22.50	19.91	22.67
Diabetes
No	88.79	63.90	90.09	98.93	< 0.0001
Yes	11.21	36.10	9.91	1.07
Sleep disorder
No	86.99	75.95	87.18	92.42	0.0006
Yes	13.01	24.05	12.82	7.58
Depression
No	92.58	86.01	92.70	95.78	0.0134
Yes	7.42	13.99	7.30	4.22
Comorbid
No	74.14	50.60	73.28	88.70	< 0.0001
Yes	25.86	49.40	26.72	11.30
BMI	29.55 ± 6.91	36.19 ± 7.34	30.01 ± 6.23	24.95 ± 4.51	< 0.0001
Cap	261.68 ± 61.39	314.67 ± 54.83	267.92 ± 55.37	218.67 ± 49.18	< 0.0001
GGT (IU/L)	20.00 (13.00, 31.00)	27.00 (19.00, 44.00)	21.00 (14.00, 34.00)	15.00 (12.00, 21.00)	< 0.0001
CRP (mg/L)	1.70 (0.81, 3.86)	3.70 (1.69, 7.33)	1.90 (0.94, 3.85)	0.970 (0.56, 1.93)	< 0.0001

LE8: Life’s Essential 8; BMI: Body mass index; GGT: Gamma-glutamyl transferase; CRP: C-reactive protein.

Furthermore, lifestyle variables, medical conditions, and biomarkers, such as marital status, race, smoking status, alcohol intake, diabetes (P < 0.0001), sleep disorders (P = 0.0006), depression (P = 0.0134), comorbidities (P < 0.0001), BMI, CAP, GGT (P < 0.0001), and CRP (P < 0.0001), were shown to be significantly correlated with LE8 Levels. These results emphasize the clear relationship between LE8 scores, clinical factors, and general health.

Association between LE8, LS7, and MAFLD

Several logistic regression models were used to analyze the effect of the main predictor variables on the result. In Model 0, the odds ratio (OR) for LE8 (per 10 units increase) was 0.51 (95%CI: 0.46-0.56, P < 0.001). In Model 1, after adjusting for age and sex, the OR remained significant at 0.53 (95%CI: 0.48-0.58, P < 0.001). An identical effect was demonstrated by Model 2, which included adjustments for drinking status, marital status, PIR, and ethnicity (95%CI: 0.43-0.59, P < 0.001). A little reduced but significant OR of 0.51 (95%CI: 0.41-0.64, P = 0.006) was obtained with Model 3, which included extra adjustments for comorbid conditions (Table 2). With regard to LE8, the unadjusted model (Model 0) showed a notable decrease in the risk of MAFLD for the moderate CVH group (OR: 0.28, 95%CI: 0.19–0.42, P < 0.001) and the high CVH group (OR: 0.05, 95%CI: 0.03–0.07, P < 0.001, P for trend < 0.001). With ORs of 0.30 (95%CI: 0.20–0.45, P < 0.001) for the moderate CVH group and 0.06 (95%CI: 0.04–0.09, P < 0.001) for the high CVH group, after additional adjustment for demographic covariates in Model 1, these associations remained strong. In Model 2, which included more socioeconomic factors, the risk reduction persisted, with the high CVH group showing an OR of 0.05 (95%CI: 0.02–0.14, P = 0.006) and the moderate CVH group showing an OR of 0.27 (95%CI: 0.13–0.59, P = 0.018). Finally, in the fully adjusted model (Model 3), which included comorbid conditions, the associations were still significant (Table 2) but moderately attenuated (moderate CVH group: OR: 0.29, 95%CI: 0.13–0.65, P = 0.022; high CVH group: OR: 0.06, 95%CI: 0.02–0.15, P = 0.007, P for trend = 0.001).

Table 2 Association of Life’s Essential 8 and Life’s Simple 7 with metabolic-associated fatty liver disease (weighted)¹.

		Model 0	P value	Model 1	P value	Model 2	P value	Model 3	P value
		OR (95%CI)		OR (95%CI)		OR (95%CI)		OR (95%CI)
LE8 (per 10)	Continuous	0.51 (0.46-0.56)	< 0.001	0.53 (0.48-0.58)	< 0.001	0.5 (0.43-0.59)	< 0.001	0.51 (0.41-0.64)	0.006
LE8 group
	Low	Reference		Reference		Reference		Reference
	Moderate	0.28 (0.19-0.42)	< 0.001	0.30 (0.20-0.45)	< 0.001	0.27 (0.13-0.59)	0.018	0.29 (0.13-0.65)	0.022
	High	0.05 (0.03-0.07)	< 0.001	0.06 (0.04-0.09)	< 0.001	0.05 (0.02-0.14)	0.006	0.06 (0.02-0.15)	0.007
	P for trend		< 0.001		< 0.001		< 0.001		0.001
LS7	Continuous	0.62 (0.57-0.68)	< 0.001	0.63 (0.58-0.68)	< 0.001	0.61 (0.53-0.70)	0.001	0.62 (0.50-0.76)	0.01
LS7 group
	Inadequate	Reference		Reference		Reference		Reference
	Average	0.31 (0.22-0.44)	< 0.001	0.32 (0.22-0.45)	< 0.001	0.29 (0.14-0.60)	0.018	0.32 (0.14- 0.74)	0.028
	Optimal	0.04 (0.02-0.09)	< 0.001	0.05 (0.03-0.10)	< 0.001	0.05 (0.01, 0.18)	0.011	0.05 (0.01- 0.22)	0.013
	P for trend		< 0.001		< 0.001		0.003		0.004

¹Model 0 Unadjusted. Model 1 Adjusted for age and sex.

Model 2 Further adjusted for ethnicity, education, marital status, poverty-income ratio, and alcohol status. Model 3 Further adjusted for comorbid conditions (diabetes, kidney failure, kidney stones, heart failure, stroke, hepatopathy, rheumatoid arthritis, and cancer).

LE8: Life’s Essential 8; LS7: Life’s Simple 7.

For LS7 (continuous), the OR in Model 0 was 0.62 (95%CI: 0.57-0.68, P < 0.001). The OR stayed significant at 0.63 (95%CI: 0.58-0.68, P < 0.001) and 0.61 (95%CI: 0.53-0.70, P = 0.001) after adjustments in Models 1 and 2. The OR in Model 3 decreased somewhat to 0.62 (95%CI: 0.50-0.76, P = 0.01). In the unadjusted model (Model 0), for LS7, the optimal CVH group showed a more marked decline in MAFLD (OR: 0.04, 95%CI: 0.02–0.09, P < 0.001, P for trend < 0.001) compared to the average CVH group (OR: 0.31, 95%CI: 0.22–0.44, P < 0.001). The ORs for the average CVH group were 0.32 (95%CI: 0.22–0.45, P < 0.001) after demographic factor adjustments in Model 1, and those for the ideal CVH group were 0.05 (95%CI: 0.03–0.10, P < 0.001). Following additional adjustments for socioeconomic status, the ORs for the average CVH group in Model 2 were 0.29 (95%CI: 0.14–0.60, P = 0.018) and those for the ideal CVH group were 0.05 (95%CI: 0.01–0.18, P = 0.011). In the fully adjusted model (Model 3), these relationships remained significant, with ORs for the average CVH group at 0.32 (95%CI: 0.14–0.74, P = 0.028) and the optimal CVH group at 0.05 (95%CI: 0.01–0.22, P = 0.013, P for trend = 0.004) (Table 2).

A nonlinear relationship was observed between LE8 and the risk of MAFLD (P for overall < 0.001, nonlinear P = 0.018) (Figure 2A). This shows a strong statistical correlation between the risk of MAFLD and LE8 scores, with considerable variations in risk as LE8 scores increase. The relationship for LS7 remained curvilinear, as seen by the nonlinearity (P for overall < 0.001, nonlinear P = 0.01) (Figure 2B).

Figure 2 Curve fitting models for Life’s Essential 8 and Life’s Simple 7 in relation to metabolic-associated fatty liver disease risk (weighted).

When weights were not taken into account, LE8 (0.740; 95%CI: 0.719-0.760; P = 0.06) and LS7 (0.752; 95%CI: 0.732-0.771; P = 0.06) had noticeably less area under the receiver operating characteristic curves (AUC) for predicting a certain condition. Nevertheless, the observed difference was not considered statistically significant (P DeLong = 0.06; Figure 3).

Figure 3  Receiver operating characteristic curve without accounting for weights.

Associations between individual CVH metrics and MAFLD

Table 3 shows how each CVH component relates to the risk of MAFLD. Higher scores in health factors (BMI, high-density lipoprotein, BP, and glucose) and health behaviors (PA) were significantly associated with a reduced risk of MAFLD (P < 0.05 for all). However, no statistically significant relationships were found between MAFLD and scores for diet (DASH), smoking, or sleep.

Table 3 Association of individual Life’s Essential 8 metric with metabolic-associated fatty liver disease¹.

CVH component (per 10)	Crude model			Adjusted model
	OR	95%CI	P value	OR	95%CI	P value
LE8 health behaviors	0.84	0.79-0.90	< 0.001	0.85	0.74-0.99	0.042
LE8 health factors	0.5	0.46-0.55	< 0.001	0.5	0.42-0.60	0.004
DASH score	0.95	0.91-0.99	0.018	0.95	0.87-1.04	0.128
PA score	0.92	0.89-0.94	< 0.001	0.92	0.88-0.97	0.023
Smoking score	0.99	0.97-1.01	0.15	0.99	0.93-1.06	0.636
Sleeping score	0.98	0.93-1.04	0.471	0.98	0.88-1.11	0.623
BMI score	0.7	0.66-0.74	< 0.001	0.69	0.61-0.77	0.006
HDL score	0.86	0.83-0.89	< 0.001	0.88	0.81-0.96	0.026
Glu score	0.75	0.71-0.80	< 0.001	0.79	0.68-0.91	0.019
BP score	0.83	0.79-0.88	< 0.001	0.86	0.77-0.97	0.034

¹Adjusted for age, sex, ethnicity, education, marital status, poverty-to-income ratio, alcohol status, diabetes, kidney failure, kidney stones, heart failure, stroke, hepatopathy, rheumatoid arthritis, and cancer.

CVH: Cardiovascular health; OR: Odds ratio; LE8: Life’s Essential 8; DASH: Dietary Approaches to Stop Hypertension; BMI: Body mass index; HDL: High-density lipoprotein; BP: Blood pressure; Glu: Glucose.

These findings highlight the complexity of the relationship between MAFLD and CVH components. More research is needed to completely understand how these components relate to mental health because some components are clearly associated with the risk of MAFLD while others are not.

Subgroup analysis

A subgroup analysis of the intervention’s impact, stratified by various demographic and clinical factors, revealed a significant positive effect in individuals aged < 40 years (OR = 0.57, 95%CI: 0.41-0.81) and a potentially significant effect in those aged 40-64 years (OR = 0.74, 95%CI: 0.58-0.95). Significant effects were also observed in both males (OR = 0.67, 95%CI: 0.50-0.89) and females (OR = 0.73, 95%CI: 0.58-0.92), with no significant gender difference (P = 0.138). Moreover, individuals with education beyond high school (OR = 0.68, 95%CI: 0.56-0.82), BMI < 25 kg/m² (OR = 0.63, 95%CI: 0.45-0.89), and high PIR (OR = 0.58, 95%CI: 0.43-0.79) experienced significant positive effects from the intervention. Although no significant differences were observed in subgroups based on sleep disorders, depression, or comorbid conditions, trends toward significance were noted for depression (P = 0.362) and comorbid conditions (P = 0.262). The consistent findings from our subgroup analyses, stratified across various demographic and clinical factors, provide further evidence for the robustness of our results (Figure 4). Similarly, in the subgroup analysis of LS7, the outcomes were robust for both the association analysis and the outcome analysis (Supplementary Figure 1).

Figure 4  Forest plot of subgroup analysis and interaction tests for Life’s Essential 8 categories in relation to the risk of metabolic-associated fatty liver disease.

Sensitivity analysis

Sensitivity analyses further corroborated the robustness of our findings, as detailed in Supplementary Tables 1 to 3. Initially, in the unweighted data, the results were consistent with those from the weighted data, indicating the robustness of the results. Additionally, within the unweighted data, curve analysis and inflection point analysis revealed the nonlinear relationship between LE8 and LS7 with MAFLD, as well as the possible threshold points (Supplementary Figures 2 and 3). These analyses utilized weighted data and incorporated alternative definitions of outcome variables, including NAFLD, SLD, and MAFLD. Consistently, our results remained stable and reliable across these different definitions (Supplementary Table 4).

DISCUSSION

Following the AHA’s 2022 upgrade of “Life’s Simple 7” to “Life’s Essential 8,” this study is the first to compare the relationships of both LE8 and LS7 CVH measures with MAFLD. Using data representative of over 120 million individuals in the United States, our study revealed that both LE8 and LS7 scores were significantly and inversely associated with MAFLD risk. In particular, a 51% decrease in MAFLD risk was linked to each 10-unit rise in LE8 score (OR: 0.51, 95%CI: 0.46-0.56, P < 0.001) according to the unadjusted Model 0. The association remained significant after adjusting for age and sex in Model 1 (OR: 0.53, 95%CI: 0.48-0.58, P < 0.001). Similar effects were produced by further adjustments in Model 2 for drinking status, marital status, PIR, and ethnicity (OR: 0.50, 95%CI: 0.43-0.59, P < 0.001). Although somewhat reduced, the correlation was still statistically significant (OR: 0.51, 95%CI: 0.41-0.64, P = 0.006) even after comorbidities were included in Model 3. A comparable pattern was observed in the continuous variable analysis of LS7. In the unadjusted Model 0, the OR was 0.62 (95%CI: 0.57-0.68, P < 0.001). Following corrections in Models 1 and 2 (OR: 0.63, 95%CI: 0.58-0.68, P < 0.001 and OR: 0.61, 95%CI: 0.53-0.70, P = 0.001, respectively), the association remained robust. A little reduced but yet significant OR of 0.62 (95%CI: 0.50-0.76, P = 0.01) was shown by Model 3, which included all modifications. Without taking weights into account, Figure 2A displays the ROC curves for LE8, with an AUC of 0.740 (95%CI: 0.719-0.760, P = 0.06) for LE8 and 0.752 (95%CI: 0.732-0.771, P = 0.06) for LS7. However, this AUC difference was not statistically significant (P DeLong = 0.06).

A statistically significant connection was identified between greater LE8 scores and decreased MAFLD risk. Significant risk reductions were observed in the moderate and high CVH groups compared to the low CVH group, with ORs ranging from 0.28 to 0.05 across the models. These associations remained strong after accounting for comorbidities, socioeconomic status, and demographic factors. Similarly, for LS7, the ideal and moderate CVH groups reported lower MAFLD risks compared to the insufficient CVH group, with ORs ranging from 0.31 to 0.04, showing a clear inverse correlation between LS7 scores and MAFLD risk. These results underline the probable mental and physical health benefits associated with better compliance with LS7 and LE8 recommendations. Higher LE8 scores were associated with a decreased risk in all three types of fatty liver disease: for each 10-unit rise in LE8 score in the fully adjusted models, the likelihood of SLD decreased by 50%, MESLD by 56%, and NAFLD by 49%. For LS7, similar trends were identified; greater scores were always associated with a reduced risk for each condition. The dose-response relationship was further highlighted by categorical analyses of LE8 and LS7. For all three outcomes, the moderate and high adherence groups had far lower risks compared to the low adherence group. The benefits of obtaining optimal CVH scores were underscored by the fact that the high adherence groups constantly demonstrated the greatest risk reduction. The robustness and generalizability of our results are bolstered by the continuous inverse associations seen across numerous outcome variables. These results suggest that the protective effects of LE8 and LS7 extend beyond a single diagnosis of fatty liver disease to encompass a wide range of metabolic health conditions and liver fat-related disorders.

Our analysis demonstrated a significant association between higher LE8 and LS7 scores and a reduced risk of MAFLD, even after adjusting for various demographic and clinical confounders. This finding aligns with previous research highlighting the protective effect of a healthy lifestyle on the development of MAFLD[25,26]. The restricted cubic spline analysis further emphasized this relationship, revealing a nonlinear, inverse association where higher scores on both scales corresponded to a significantly lower MAFLD risk. Interestingly, examining individual LE8 components revealed a nuanced interplay between health behaviors and factors in relation to MAFLD. While higher scores in health behaviors such as PA and BMI were unexpectedly associated with increased MAFLD risk, higher scores in health factors such as high-density lipoprotein cholesterol and glucose levels were associated with a decreased risk. This seemingly paradoxical finding regarding health behaviors might be attributed to complex interactions between lifestyle factors and development of MAFLD, potentially influenced by factors such as dietary habits not captured in the LE8 scoring system. Further investigation is warranted to clarify these relationships[27-30].

Although both LE8 and LS7 demonstrated significant associations with MAFLD risk, no statistically significant difference was observed in their predictive ability for MAFLD, as indicated by the AUC analysis. This suggests that both metrics can be valuable tools for assessing MAFLD risk in clinical settings. However, the choice between LE8 and LS7 may depend on the specific context and the availability of data, considering LE8 incorporates sleep health, which might be relevant for specific populations, while LS7 was designed to provide simpler and more actionable health improvement goals[31-33].

PA emerged as the strongest correlate, consistent with the findings of previous studies[34]. Huang et al found that active leisure-time PA is associated with a reduced risk of MAFLD/MASLD and liver fibrosis. Importantly, this association was not observed for occupational or transportation-related PA, highlighting the specific benefits of leisure-time PA for liver health[35]. This study analyzed the 2017-2018 NHANES dataset and found that both active PA (especially leisure-time PA) and adequate weekday sleep duration were independently associated with a lower MASLD risk. The combination of these lifestyle factors further reduced the risk, except among those with significant alcohol consumption[36]. A case-control study investigated micro-RNAs (miRNAs) as potential biomarkers for SLD, a condition often linked to a high-fat diet and lack of exercise. While individual miRNA levels showed some correlation with disease, combinations of miRNA changes demonstrated stronger associations with disease severity, suggesting potential diagnostic and prognostic applications[37]. Previous studies exploring individual dietary metrics in LS7 showed that the association between diet and MAFLD remains controversial after adjusting for other LS7 metrics[38].

CONCLUSION

This research examined the LE8 and LS7 scoring systems for predicting the risk of MAFLD. The LE8 score system covers more healthy living characteristics, including sleep health, while the LS7 system focuses on fundamental health behavior markers. This research found that both scoring systems predict MAFLD risk with similar accuracy despite their design variations. In practice, either the LE8 or LS7 scoring system can be used to evaluate and manage MAFLD risk, depending on personal circumstances and preferences. Regardless of the scoring method used, adopting a healthy lifestyle—especially maintaining a healthy BMI and engaging in regular exercise—is essential to prevent MAFLD. Although a balanced diet and smoking cessation are important for overall health, this research indicates that they have a comparatively smaller impact on reducing MAFLD risk. Healthcare practitioners should prioritize promoting and personalizing lifestyle interventions to assess individual needs. However, the intricate relationship between CVH factors and MAFLD warrants further investigation to deepen our understanding and optimize prevention strategies.

ACKNOWLEDGEMENTS

We are deeply grateful to Jie Liu from the General Hospital of the People’s Liberation Army (301 Hospital) for her valuable support in clinical research methods and ideas.